Tanzania Journal of Agricultural Sciences1998, Vol.1(2)'; 131-140Several Bean growing areas in Tqnzania have unreliable and marginal rains necessitating use of short maturing and high yielding varieties. Such varieties however, are not, readily available. This'study was done to evaluate varieties for maturity, yield and its components in the bean growing areas of Morogoro, Tanzania. Sixty four varieties from the Sokoine University of Agriculture (SUA) collection and from a high percentage protein population were grown at two locations, Mgeta (high altitude) and SUA (low altitude) in 8 x 8 partial lattice with 3 replicatiqns during the 1991 season. Yield levels were low, ranging between 0.2 - 0.8 tlha at Mgeta and 0.06 - 2.Otlha atSUA. Earlier varieties were superior in yield due to' a higher yield efficiency. Varieties grown in the cooler higher altitude areas matured late with consequent reduction in yield. Significant G x E interaction effects were observedfor all the variables, except number of seeds per pod. Positive genetic correlations were significant (P ~0.01) at both locations between yield and number of pods per plant, yield efficiency and podjilling period. Yield was genetically and negatively corrected with days to first flower, 50% podfill and 85 % maturity. Path coefficient showed that days to first flower had a consistent negative direct effect o!'l yield. At both locations, the negative genetic relationship between days to 50% podfill with yield was attributed to the negative indirect influence through days to first flower. Earlier flowering anq maturity, longer podfilling with high pod production and yield efficiency are recomme.nded for higher yields of beans in the study area.